Back to skill

Security audit

lifelog2

Security checks for vulnerabilities and agentic risk

Overview

This diary skill stores entries locally and can archive them to flomo and ima when asked; the sensitive behavior is disclosed and user-directed.

Install only if you are comfortable with diary entries being saved in local draft files and, when you run the archive command, being sent to both flomo and ima. Review the summary before archiving, and avoid recording secrets or highly sensitive personal details unless those destinations are acceptable to you.

Vulnerability Patterns
  • Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
  • Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
  • Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
  • Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally
  • Unauthorized Access and Privilege EscalationObtains permissions beyond the task's legitimate needs
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (13)

Missing User Warnings

High
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill sends private diary data to external services (flomo and ima) during archiving, but it does not require explicit informed consent at the point of transfer or provide a privacy warning in the description. Exfiltrating intimate user content to third parties can significantly increase exposure, especially if those services retain, analyze, or further process the data.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
90% confidence
Finding

The instructions require a specific Chinese locale format such as YYYY 年 M 月 D 日, and the examples and responses are fixed in Chinese. This is a language/locale constraint presented as mandatory behavior, with no user opt-in or alternative locale choice documented.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The skill persistently stores highly sensitive diary content in a local file path under the user's home directory, but it does not clearly warn users that their private entries will be retained on disk across sessions. This creates privacy and confidentiality risk because local files may be readable by other local processes, backups, sync tools, or other users depending on system configuration.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The eval uses a very broad natural-language trigger phrase, "记日记", without any boundary conditions or disambiguation rules. In an agent skill, this can cause unintended activation when users mention the phrase incidentally, leading to accidental capture of private diary content or invocation of downstream journaling behavior.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
88% confidence
Finding

The summarize/archive flow is triggered by underspecified phrases like "汇总日记" that do not constrain scope, target dataset, or user confirmation. This increases the risk of summarizing the wrong entries or performing a sensitive operation on unintended content, especially in a skill that handles personal notes and stateful temporary storage.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
91% confidence
Finding

The archiving trigger "日记存档" is vague while the expected behavior includes writing data to external systems (flomo and ima) and clearing temporary state. Because this is a destructive and externally side-effecting action, ambiguous activation can lead to accidental exfiltration of personal diary content and unintended deletion of local temporary records.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
96% confidence
Finding

The archive workflow explicitly deletes the local draft diary file after sending content to external destinations, but the documented flow does not mention any confirmation, warning, backup, or recovery option. For a journaling skill handling personal data, silent deletion can cause irreversible user data loss if archival fails partially, the external write is incorrect, or the user did not intend destructive cleanup.

Content

No source excerpt is available for this finding.

Missing User Warnings

Medium
Category
Not specified by scanner
Confidence
97% confidence
Finding

The skill describes sending diary contents to external skills/services (flomo-notes and ima-skill) without any privacy notice, consent step, or explanation of where sensitive personal text will be stored. Because diary entries commonly contain intimate thoughts, schedules, and personally identifiable information, undisclosed transmission to third-party destinations materially increases privacy and confidentiality risk.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Medium
Category
Not specified by scanner
Confidence
95% confidence
Finding

The document declares lang="zh-CN" and all visible UI text and usage examples are presented only in Chinese. Under the policy, forcing a specific language without user opt-in is a natural-language policy violation unless the locale restriction is explicitly justified or alternatives are offered.

Content

No source excerpt is available for this finding.

Vague Triggers

Medium
Category
Not specified by scanner
Confidence
89% confidence
Finding

The test document uses very broad trigger phrases such as “记日记”, “汇总日记”, and “日记存档” without showing any scope constraints, confirmation requirements, or guardrails around when these actions should execute. In a skill that records and archives personal content and can invoke other skills, overly broad invocation phrases increase the chance of accidental activation, unintended data capture, or unreviewed cross-skill actions.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
87% confidence
Finding

The document states that the title format must be YYYY 年 M 月 D 日#lifelog, which hardcodes a Chinese locale-specific output convention. Because no choice or justification is provided, this constitutes a natural-language locale policy concern.

Content

No source excerpt is available for this finding.

Natural-Language Policy Violations

Low
Category
Not specified by scanner
Confidence
84% confidence
Finding

The file's user-facing examples and operational text are exclusively in Chinese, which can amount to a language policy issue when no user choice or locale justification is provided. There is no indication that the skill offers multilingual support or explicitly limits itself to a Chinese-speaking context.

Content

No source excerpt is available for this finding.

Missing User Warnings

Low
Category
Not specified by scanner
Confidence
93% confidence
Finding

This HTML file includes JavaScript that collects textarea contents and triggers a download of feedback.json, which is a user-data write operation. There is only a success alert after the write is initiated, but no prior user-facing warning in the page description or near the button explaining that feedback will be exported to a local file.

Content

No source excerpt is available for this finding.